....Bromine released from atmospheric CH3Br is highly effective in depleting stratospheric ozone. Recent budget estimates for CH3Br suggest that sources exceed sinks by about 40 Gg y-1, demonstrating a need for additional understanding of the behavior of this gas in nature [Butler and Rodriguez, 1996], particularly its air-sea flux. Lobert et al.  reported that most of the East Pacific Ocean was undersaturated and probably a net sink for this gas. Those results were further supported by findings for the Atlantic Ocean, suggesting that the world's oceans could be a global, net sink for atmospheric CH3Br [Lobert et al., 1996].
....Two recently published, numerical models suggested that the polar oceans might be a large, net source of atmospheric CH3Br [Pilinis et al., 1996; Anbar et al., 1996]. The two models used production rates based on data published in Lobert et al. , presuming them to be either constant over the entire ocean or a function of chlorophyll-a concentration. In the latter case, production was correlated with satellite ocean color data as a proxy for chlorophyll-a concentration and those correlations were extrapolated to polar regions. With chemical degradation being largely suppressed in cold, polar waters, and a very high biological productivity during the austral summer, the predicted saturation anomalies were positive and ranged up to 500%, indicating that this polar source could globally outweigh the sinks estimated in Lobert et al. . To resolve this discrepancy, our study was conducted to measure the actual saturation of CH3Br in the Southern Ocean during a time of high biological productivity.
....The BLAST III cruise was conducted between 21 February and 07 April, 1996, started in McMurdo, Antarctica (78°S), and ended in Punta Arenas, Chile (54°S; Figure 1).
....The gas chromatograph / mass spectrometer (GC/MS) that was used to measure about 15 compounds including CH3Br was virtually identical to the system used during the two previous cruises [Lobert et al., 1995; 1996] with slight modifications in the trapping procedure. On this cruise we also measured CH3Br with a custom-built GC equipped with an electron capture detector (GC/ECD) and different columns and ECDs in support of measurements by GC/MS. The sampling system and the equilibrator were identical to those used on previous cruises. Measurements by both GC/MS and GC/ECD were calibrated with the same whole-air standards, which were calibrated against gravimetrically prepared standards. Chlorophyll-a fluorescence was measured in surface waters with a continuous-flow fluorometer (Turner Designs, Model 10), which was calibrated with chlorophyll standards following the cruise.
....Saturation anomalies, corrections for physical effects, and air-sea fluxes were derived following the approach of Butler et al.  with the solubility, degradation rate, Schmidt number, and wind speed correction as described in Lobert et al. .
Results and Discussion
....Measured, dry mole fractions of CH3Br in the atmosphere averaged 8.3 ± 0.3 ppt (GC/MS) and 8.5 ± 0.7 ppt (GC/ECD) over the entire cruise (Figure 2a) and are consistent with data from the BLAST I and BLAST II cruises. CH3Br was undersaturated in the water, with a mean dry mole fraction in the equilibrated air of 5.5 ± 0.6 ppt (GC/MS) or 5.6 ± 0.8 ppt (GC/ECD). Mole fractions from MS and ECD systems agreed, on average, within 0.2 ppt, showed identical substructures everywhere, and were correlated at >>99.9% confidence. Most important, however, is that the ocean was consistently undersaturated in CH3Br, with a mean saturation anomaly of 36±7%, which becomes -33±8% after correcting for physical effects such as the seasonal warming of the surface ocean [Butler et al., 1991]. Similarly large undersaturations were recently found by Moore and Webb  for the Labrador Sea in the summer.
....The calculated, chemical degradation rate for CH3Br in these cold waters is 0.6% d-1 (D.B. King and E.S. Saltzman, Removal of methyl bromide in coastal seawater: Chemical and biological rates, submitted, J. Geophys. Res., 1996). Maintaining a steady-state, 35% undersaturation of CH3Br in the surface waters in the presence of air-sea exchange and with the absence of any in situ production requires an in situ degradation rate of about 5.8 % d1 at the given temperatures and wind speeds, which is a factor of 10 larger than that for chemical degradation alone. The most likely explanation of these findings is that dissolved CH3Br is being degraded by an additional, significant mechanism other than reaction with H2O and Cl-. In fact, current laboratory research suggests that biological removal of CH3Br in subtropical waters is about as fast as the chemical sink (King and Saltzman, 1996).
....Thus, if there is any in situ production of CH3Br in the Southern Ocean, then the degradation rate must be even higher than calculated in order to maintain the observed saturation anomaly. Our earlier estimates for aquatic production of CH3Br were based upon the difference between net flux and chemical loss [Lobert et al., 1995; 1996], but they did not include consideration of an additional sink. If this additional loss mechanism is significant and ubiquitous, then oceanic production of CH3Br must be higher than previously estimated.
....Our shipboard measurements show that the concentration of dissolved CH3Br decreased along with chlorophyll-a concentration (between 0.8 and 14 µg chlorophyll-a l-1) during the cruise. This contrasts with findings of Moore and Webb , who showed that chlorophyll-a above 1 µg l-1 correlated negatively with CH3Br concentration in Labrador Sea and Northern Gulf Stream waters. Taken together, these results suggest that a simple, universal relationship between chlorophyll-a and dissolved CH3Br does not exist. Because the large supersaturations suggested by models of Pilinis et al.  and Anbar et al.  are not supported by our observations, it becomes clear that the saturation or the net, air-sea flux of CH3Br cannot be estimated directly from chlorophyll concentration or ocean color. Aquatic production could still be high and may even be correlated to chlorophyll, but it cannot be deduced from our measurements and must be balanced by equally high degradation.
....Finally, from the corrected saturation anomaly of
CH3Br, we estimate a net air-to-sea flux of 8.0 Gg
y-1 for the polar oceans. This estimate includes assumptions
that the undersaturation is independent of the season, that the
investigated area represents polar oceans in general, and that
polar oceans constitute 10% of the whole ocean. Under these conditions,
our previously published estimates of the global, net air-sea
flux of atmospheric CH3Br [Lobert et al., 1995;
1996] become more negative, with a best estimate of 21 (-11 to -32) Gg y-1. The overall uncertainty of 50% is mostly due to uncertainties in the air-sea exchange calculation. The
data presented here place considerable constraints upon CH3Br
emissions from polar waters, and provide further evidence for
a global, net, oceanic sink for atmospheric CH3Br.
.... Acknowledgments. We thank Antarctic Support Associates and the crew and captain of the R/V Nathaniel B. Palmer for their support and we are indebted to Francisco Chavez for calibrating our chlorophyll measurements. This work was funded by the Methyl Bromide Global Coalition, the Atmospheric Chemistry Project of NOAA's Climate and Global Change Research Program, and a Global Change Distinguished Postdoctoral Fellowships sponsored by the U.S. Department of Energy.
Data from this cruise are available for public use via world wide web at ftp://ftp.cmdl.noaa.gov/noah/ocean/blast_iii/.